ACT Science Research Summaries: Experiments, Procedures, Results, and Follow-Up Trials

Research Summaries are the most common ACT Science passage format, currently about 45–60% of Science passages.

They present:

• one or more experiments;
• procedures;
• variables;
• figures/tables;
• results;
• sometimes hypotheses.

The biggest mistake is reading every procedural sentence with equal attention.

The better method is:

map experiment structure first, then retrieve details question-by-question

What to map

For each study, identify:

• goal;
• independent variable;
• dependent variable;
• control/baseline;
• key constants;
• results.

A five-line map can replace rereading a long passage.

Example experiment map

Study 1:

• change: temperature;
• measure: reaction rate;
• constants: pH, concentration;
• result: rate rises to 30°C, then falls.

Study 2:

• change: pH;
• measure: reaction rate;
• constant: 30°C;
• result: peak at pH 7.

Now many questions become lookup/comparison.

Read questions before deep procedure

A useful Science strategy:

1. skim passage organization;
2. inspect figures;
3. read question;
4. retrieve exact detail.

You often do not need to memorize apparatus dimensions or reagent names.

Study labels matter

If question says:

Based on Study 2...

do not mix Study 1 results.

ACT distractors often use correct data from wrong study.

Procedure vs result

Procedure:

what researchers did.

Result:

what happened.

Question:

"Which variable was held constant?"

Look in procedure.

Question:

"At 20°C, which condition had greatest rate?"

Look in results/figure.

Know where information lives.

Compare studies

Common question:

How did Study 2 differ from Study 1?

Make a difference table:

Then answer.

Same result, different design

Two studies can measure same dependent variable under different independent variables.

Do not assume same outcome means same procedure.

Hypothesis questions

Passage may state:

Researchers predicted X increases Y.

Results can:

• support;
• partially support;
• contradict.

Use exact pattern.

If Y rises then falls, answer:

hypothesis supported only over part of range

rather than:

fully supported.

Control group questions

If control receives:

• no treatment;
• standard treatment;
• baseline condition,

use it to compare treatment effect.

The control is not necessarily:

lowest numeric value.

It is defined by experimental role.

Follow-up trial questions

Question:

If experiment repeated at 40°C, which outcome is most likely?

Use trend from observed data.

This may require:

• interpolation;
• extrapolation;
• mechanistic statement given in passage.

Do not rely on outside science unless necessary.

Additional-trial design

Question:

Which experiment best tests whether light intensity caused effect?

Choose design where:

• intensity changes;
• wavelength and other factors fixed;
• same outcome measured.

See ACT Science Variables and Controls.

Procedure order

Some questions ask:

What happened after step X?

Use passage sequence.

Do not reorder based on what would make scientific sense.

The passage is source.

Apparatus diagrams

A figure may show:

• sensor placement;
• layers;
• flow direction;
• sample position.

Question can ask:

Which component measures temperature?

Read labels.

Do not ignore diagrams because they look technical.

Tables across experiments

If table columns correspond to Study 1 and Study 2, confirm:

• same units?
• same variable?
• same conditions?

Avoid cross-column miscomparison.

Normalize experiment names

Long labels can be reduced.

"Solution containing 0.10 M NaCl at 25°C" becomes:

low salt / 25°C.

This reduces cognitive load.

Worked example 1: compare variables

Study 1 varies:

concentration.

Study 2 varies:

temperature.

Both measure:

reaction time.

Question:

What variable differs in purpose between studies?

Answer:

independent variable changed from concentration to temperature.

Worked example 2: identify constant

All trials:

• same mass sample;
• different pH;
• same time.

Constant:

mass and time.

Do not choose:

pH.

Worked example 3: hypothesis support

Predicted:

higher concentration → higher rate.

Data:

• 1 M: 5
• 2 M: 8
• 3 M: 8
• 4 M: 7

Best:

supported from 1 to 2 M, but not continuously across all concentrations.

Worked example 4: follow-up

If rate peaks at 30°C and declines by 40°C, likely value at 35°C:

between 30 and 40°C values, if trend smooth.

That is interpolation.

Experimental replication

If multiple trials give similar pattern:

evidence more reproducible.

If trials vary widely:

more uncertainty.

Do not claim:

one consistent replication proves universal law.

Experimental controls and causal claims

A controlled experiment can support causation if:

• target variable manipulated;
• alternatives controlled;
• outcome differs systematically.

But answer should match passage's evidence strength.

Research Summary reading order

First 20 seconds

Read:

• study titles;
• what changed;
• what measured;
• figure axes.

Then

Questions drive detail retrieval.

This is often faster than reading every line first.

When full passage reading is useful

Read more deeply if:

• question asks overall purpose;
• procedure is complex;
• studies differ subtly;
• prediction depends on mechanism.

Strategy should adapt.

Multiple experiments and nested controls

Study 1 may establish baseline.

Study 2 adds variable.

Study 3 changes second factor.

Create compact matrix:

This makes interactions visible.

Engineering research summaries

Current ACT can include design constraints.

Example:

• material strength;
• cost;
• weight.

Study compares designs.

Question may ask:

Which design meets both strength and cost criteria?

Treat like multi-variable table.

Common traps

Trap 1

Right result, wrong study.

Trap 2

Procedure confused with result.

Trap 3

Independent/dependent reversed.

Trap 4

Hypothesis called fully supported when only partial.

Trap 5

Extrapolating too far.

Trap 6

Using outside science instead of passage.

Trap 7

Ignoring units.

Fast question-family routes

Direct result

Go figure/table.

Variable/control

Go procedure.

Compare studies

Make difference list.

Follow-up

Use trend/design.

Hypothesis

Compare prediction vs results.

New experiment

Hold controls, vary target.

Current ACT Science context

Research Summaries currently make up about 45–60% of ACT Science passage formats. Scientific Investigation is about 18–32% of Science questions, while Interpretation of Data is roughly 38–50%.

So Research Summaries often combine experimental design and graph reading in the same passage.

Practice mini-map

Study:

Seeds grown at 10, 20, 30°C. Germination measured after 7 days.

Independent:

temperature.

Dependent:

germination.

Constant:

seed type, water, duration.

If highest germination at 20°C:

data suggest tested optimum near 20°C, not that 20°C is universal optimum for all seeds.

How to review mistakes

Classify:

• wrong study;
• procedure/result confusion;
• variable;
• control;
• trend;
• interpolation/extrapolation;
• hypothesis strength;
• outside knowledge.

Build a study matrix for multi-experiment passages

When three experiments differ in several ways, use a compact matrix.

This makes comparison questions almost mechanical.

Distinguish replicated experiment from new experiment

If Study 2 repeats Study 1 under same conditions with new samples:

replication.

If it changes a major factor:

extension/new experiment.

Question may ask why Study 2 was performed.

Replication checks:

consistency.

Extension tests:

generalization or new variable.

Predicting a missing table value

If table follows a stable pattern and one value is omitted, infer only if relationship supports it.

Example:

• concentration 1 → rate 4
• concentration 2 → rate 8
• concentration 3 → ?

If passage/graph indicates direct proportionality:

12.

If no such relationship is established:

do not assume.

Compare procedures, not just results

Two experiments may reach similar results through different procedures.

Question:

Which step occurred in Study 1 but not Study 2?

Scan procedure list, not graph.

Procedure-comparison questions reward organization more than science knowledge.

Mechanism questions

Sometimes passage explains why a result occurs.

Example:

increased temperature raises molecular collision frequency.

Then a prediction question may require using that mechanism.

Distinguish:

• empirical trend only;
• stated mechanism.

If mechanism is given, it can justify stronger extrapolation.

Null result

If treatment and control are nearly identical, conclusion may be:

no detectable effect under tested conditions.

Do not translate this into:

treatment can never have an effect.

Scope:

• tested conditions;
• measured outcome;
• study duration.

Unexpected result

A result can contradict hypothesis.

Best interpretation:

data do not support original prediction in that trial.

Do not automatically conclude:

experiment failed.

Unexpected data can still be valid evidence.

Multi-step calculation questions

Research Summaries can require:

• difference;
• ratio;
• percent change;
• average.

Write values with units before calculating.

Example:

$$\frac{12-8}{8}\times100=50\%$$

Do not confuse absolute difference 4 with percent change 50%.

Error-analysis framework

If a Research Summary question is wrong, identify whether you:

• used wrong study;
• used wrong figure;
• confused procedure/result;
• misread variable;
• assumed a trend;
• ignored units;
• overgeneralized beyond tested conditions.

That diagnosis directly informs practice.

When studies use different dependent variables

Two studies can test the same broad hypothesis but measure different outcomes.

Example:

Study 1:

plant height.

Study 2:

dry mass.

Do not assume the values are directly comparable.

The shared purpose may be:

assess growth,

but the measurements differ.

A comparison question may ask:

Which study measured a more direct structural outcome?

Answer from definitions in passage.

Baseline normalization

A study may express results as:

$$\frac{\text{treatment}}{\text{control}}$$

rather than raw values.

Interpret:

• 1.0 = same as control;

• 1 = greater;

• <1 = lower.

Do not confuse normalized values with the original units.

Unexpected procedure details

If an experiment includes a wash step, waiting period, or calibration, ask whether the question targets:

• why it was done;
• what variable it controls;
• whether it changes comparability.

These procedural details are often there because they make the design fair or measurements valid.

Reading dense methods efficiently

For a long paragraph, annotate mentally:

• sample;
• treatment;
• duration;
• measurement.

Example:

"Twenty seedlings were placed in chambers at three humidity levels for seven days. Leaf mass was then measured."

Compress to:

seedlings / humidity / 7 days / leaf mass.

That is enough for most questions.

When two studies disagree

If Study 1 finds increase but Study 2 finds no change, do not force a single combined conclusion.

Best answer may be:

results were inconsistent across studies.

Then look for design differences that could explain the discrepancy.

ACT often asks you to identify that difference rather than resolve the science.

FAQ: ACT Research Summaries

How common are Research Summaries?

They are currently the largest Science passage format, roughly 45–60%.

Should I read every procedure first?

Not necessarily. Map the design, inspect figures, and let questions direct detailed reading.

What should I write down?

Independent variable, dependent variable, control/constants, and major result pattern.

How do I answer follow-up questions?

Use the established design and data trend while preserving experimental scope.

Do I need advanced science?

No. Advanced knowledge is not required; passage/data are primary.

Related MastaPrep guides

• ACT Science Experimental Design
• ACT Science Variables and Controls
• ACT Science Graph Trends
• ACT Science Conflicting Viewpoints
• ACT Science Strategy

Final takeaway

Research Summaries reward experiment mapping:

change → measure → control → result

Do not memorize every procedural detail.

Build the structure, then retrieve exactly what each question needs.

---

Official ACT references:
https://www.act.org/content/act/en/products-and-services/the-act/test-preparation/description-of-science-test.html
https://www.act.org/content/act/en/students-and-parents/high-school-success/testing-advice-for-the-act/act-science-test-tips.html

Trademark note: ACT® is a registered trademark of ACT Education Corp. MastaPrep is independent and is not affiliated with or endorsed by ACT Education Corp.